EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Novel Feature Selection Method Based on an Improved Beluga Whale Optimization Algorithm for Thermal Comfort Prediction

C. Premila Rosy, Saravanan K, S. Josephine Theresa, D. Meenakshi, D. Geethamani

Engineering, Technology & Applied Science Research · 06.06.2026 · 16 (3) · 35811-35817

DOI: 10.48084/etasr.17550 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Engineering, Technology & Applied Science Research, online erschienen am 06.06.2026. Verfasst von C. Premila Rosy und 4 weiteren Autor:innen. Die Arbeit nennt 22 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Heating, Ventilation, and Air Conditioning (HVAC) systems consume a significant portion of building energy and do not usually address individual thermal comfort requirements. To overcome this difficulty, the present study proposes a privacy-aware, deep learning-based thermal comfort prediction system using Federated Deep Learning (FDL). The preprocessing of the ASHRAE Global Thermal Comfort Database II was performed through outlier removal, Z-score normalization, categorical encoding, and class imbalance using the Synthetic Minority Oversampling Technique (SMOTE). The Cubic Transverse Mutation Beluga Whale Optimization (CTMBWO) algorithm was employed to effectively identify features without redundancy and enhance robustness. A federated deep neural network was trained on the chosen features without learning raw data. Chicken Swarm Optimization (CSO) was applied to optimize hyperparameters and maximize convergence and accuracy. Five-fold stratified cross-validation achieved a mean of 98.14 with low variance, outperforming baseline models. The proposed model facilitates energy-saving and privacy-conscious HVAC management in intelligent buildings.

Zitieren

APA

Rosy, C. P., K, S., Theresa, S. J., Meenakshi, D., & Geethamani, D. (2026). A Novel Feature Selection Method Based on an Improved Beluga Whale Optimization Algorithm for Thermal Comfort Prediction. Engineering, Technology & Applied Science Research, 16(3), 35811-35817. https://doi.org/10.48084/etasr.17550

Vancouver

Rosy CP, K S, Theresa SJ, Meenakshi D, Geethamani D. A Novel Feature Selection Method Based on an Improved Beluga Whale Optimization Algorithm for Thermal Comfort Prediction. Engineering, Technology & Applied Science Research. 2026;16(3):35811-35817. doi:10.48084/etasr.17550

BibTeX
@article{Rosy2026,
  author = {Rosy, C. Premila and K, Saravanan and Theresa, S. Josephine and Meenakshi, D. and Geethamani, D.},
  title = {A Novel Feature Selection Method Based on an Improved Beluga Whale Optimization Algorithm for Thermal Comfort Prediction},
  journal = {Engineering, Technology & Applied Science Research},
  year = {2026},
  volume = {16},
  number = {3},
  pages = {35811--35817},
  doi = {10.48084/etasr.17550}
}
Kostenlos starten →